what and where is the obstacle that gives the 5-something % gradient for that runway?
One would need to check with the airport or the relevant surveyor to identify the specific obstruction. However, a quick looksee at the RDS data indicates that it is likely to be the boundary fence or associated lights. Simple matter to run the trig to get an approximate distance from end TODA .. but probably not necessary in this case.
wouldn't it be operating under CAO 20.7.4
Probably
would it or would it not be required to meet that 5-something % obstacle gradient?
Not at all. It is most likely that one of the earlier STODA figures would be used or, if one is a bit clever, run the sums and use an intermediate distance/gradient if the calculations are reasonably defining. In this case I would anticipate that the STODAs cross at the one critical obstacle.
A case of matching weight limitations for TODR against climb gradient capability. One would need to check the POH figures to make a more definite assessment. Not that it makes for much value as the reconfiguration puts it out of reach unless the runway has lots of spare distance ... In practical terms, a FAR23 twin doesn't have much OEI value until it is somewhat above the aerodrome and other obstacle levels ie for the typical GA aerodrome the pilot should be in silent prayer that nothing quits in the first how ever many hundred feet of climb .. Hence the oft-referred to "grey" area of the takeoff.
under 20.7.4 is that it must be able to make 6% normal ops
Correct. If that is a limit for the particular model, there will be a climb weight limit chart of some sort in the POH data.
and 1% with prop feathered and gear & flap up.
Actually, better than that .. the 1% for charter is determined at 5000 ft ISA. The gotcha, in practical terms, is the distance taken to effect the configuration and speed changes .. for the typical runway it doesn't work out and we just go on the AEO case with the OEI looking from sometime after the initial takeoff.
The OEM accel-go data, where provided, usually is sufficiently alarming ... if we were concerned about an early failure, we would go back to the office for a coffee instead of going flying.
Quite different to 20.7.1B.
Significantly so